Share this article on:


Carlson, Robert H.


Although two independent preclinical studies at academic centers have found that an amyloid beta protein vaccine restored mouse memory and problem-solving ability, the question remains whether this treatment will ever alleviate symptoms of Alzheimer disease (AD) in humans.

In phase I clinical trials of vaccine AN-1792, developed by Elan Corporation, human subjects who received single or multiple vaccinations exhibited no adverse effects from the drug, according to company reports last year. The drug was safe enough to proceed with phase II testing, but Elan officials say data are still being analyzed and they cannot say whether the human subjects showed any improvement in cognitive performance.

The preclinical data appear to support the amyloid cascade hypothesis, that amyloid plaque deposits are the cause of AD and that decreasing plaque burden relieves AD symptoms.

“Anti-amyloid treatments are predicated on the amyloid hypothesis being fundamentally sound,” Dennis J. Selkoe, MD, Professor of Neurology at Harvard Medical School and Brigham and Women's Hospital, explained in an interview. Dr. Selkoe said he has felt for many years that amyloid is the cause of AD and not an effect.

“I am sure if you clear away amyloid plaques with some kind of vaccine, people will do better neurologically,” Dr. Selkoe said. “Or if it's used in a preventive mode, they will not get symptoms in the first place.”

Back to Top | Article Outline


Dr. Selkoe and colleagues are testing an intranasal amyloid beta vaccine. Published data appeared last October in Annals of Neurology showing that it led to a very definite clearing of plaques in the brain.

“We got a strong antibody response, and all our data suggest that intranasal administration has a very similar effect to what Schenk and colleagues got by injection,” Dr. Selkoe said, referring to AN-1792's inventor, Dale Schenk, PhD, Elan's Vice President of Discovery Research. “Regardless of the route of administration, I think if you get a strong antibody response in humans, the antibodies will circulate to the brain.”

In a phone interview, Dr. Schenk said: “Less than one percent of antibodies cross the blood-brain barrier in mice, but fortunately, that seems to be enough. Our measurements in mice show the amount of antibody that move in to the brain is exactly the amount you would predict in animals, and that happens to be very similar to what you see in humans, about 0.1 to 0.3 percent.”

Back to Top | Article Outline


No test is directly comparable between animals and humans, but the senior author of one preclinical vaccine study said the Morris water maze test he used does show whether mice improve in spatial memory and solution of complex tasks.

Peter H. St. George-Hyslop, MD, Professor in the Department of Medicine at the University of Toronto and a neurobiologist in the Center for Research in Neurodegenerative Diseases there, said that in controls, the mouse understands it is in a tank of water and must swim to and climb onto a platform. “The animal develops a cognitive paradigm to do that, and it must also remember the location of the platform and rely on visual clues around the room to find that platform, so it's a combination of learned complex tasks and spatial memory,” he said.

The Canadian researchers reported late last year that mice treated with the A-beta-42 vaccine had a 50-percent reduction in the amyloid plaques and also showed improved performance in the water maze, as compared with mice given an islet associated polypeptide vaccine not associated with the CNS.

In another trial published at the same time, University of South Florida researchers saw improved working memory in an AD mouse model when challenged in a radial-arm water maze after receiving a vaccine.



Dr. St. George-Hyslop believes that animal performance in these tests directly corresponds to a healthy human who learns and uses geographical references relative to their home.

“Patients with early- and mid-stage AD have terrible problems with geographic navigation,” he said. “They cannot remember spatial coordinates or develop appropriate cognitive paradigms to find their way.”

Back to Top | Article Outline


Extrapolating from AD-like animal models is risky because humans have other aspects of pathology, and because the vaccines will differ. Both engineered mice and humans with AD have amyloid burden, plaques, dystrophic neurites, and gliosis, said Roger N. Rosenberg, MD, Chair and Professor of Neurology and Director of the Alzheimer's Disease Center at the University of Texas Southwestern Medical Center in Dallas. But the mouse model does not have neurofibrillary tangles or hyperphosphorylated tau-containing tangles within nerve cells, he said.

“One has to be careful about correlations of pathology and improvements with vaccinations in mice and whether it would follow in patients,” Dr. Rosenberg said. “But the pathology between mice and humans is close, and the animal response to a vaccine is suggestive that patients may similarly improve.”

Dr. St. George-Hyslop said his laboratory's A-beta-42 vaccine was fairly easy to synthesize, purify, and mix with an adjuvant. In his study it was first injected into mice interperitoneally, and then subcutaneously. He said it is still not known whether humans will require a series of booster immunizations for active immunization, as in the preclinical work, or whether it is possible to immunize passively with purified antibodies. He believes the latter would be safer.





“A potential adverse vaccine side effect in humans would be an immune response that also causes an auto-immune reaction that can't be switched off,” he said. “But if you give the antibody passively and you get side effects, you could just stop it and the issue would resolve.”

Dr. St. George-Hyslop said human vaccine researchers will also have to determine whether the vaccine produces a vigorous immune response in mature humans. “Their targets will be adults in their 40s, 50s, and 60s. Vaccines can produce an immune response in the older subjects – influenza vaccines, for example – but it is sometimes difficult to do in older subjects.”

A question that probably may not be answered for many years is what happens when the vaccine is given to a person in whom the disease is quite established. “Will the cognitive deficit be reversible or is it fixed?,” Dr. St. George-Hyslop speculated. “And what if you immunize and then withdraw the immunization, do the plaques come back?”

Back to Top | Article Outline


A vaccine that shows the ability to arrest the progress of the disease will almost immediately be considered as a potential preventative medication, said Bill Thies, PhD, Vice President for Medicine and Science at the Alzheimer's Association.

“Obviously there are questions about how often it has to be given and the potential hazards, but the ultimate goal here is to come up with a product that you can give to people in their 40s and prevent the disease from ever occurring,” Dr. Thies said. The long-term effects of a vaccine will not be seen until many patients have taken it for years, but so far in the animal models there has been no evidence that antibodies are binding to the amyloid precursor protein present on the surface of cells, which could cause some disturbance in cell function.



“We now have an idea that this [immune response to AN-1792] isn't some bizarre effect; it really is a typical immune stimulation of antibodies that causes the plaque clearance,” Dr. Thies said.

“What we can't do at this point is draw any conclusions about effectiveness in humans,” he continued. “Those mice don't have Alzheimer disease; they have plaque accumulation.”

AN-1792 is the only AD vaccine Dr. Thies is aware of, but he said many other companies are developing an AD treatment with a different mechanism – the inhibition of secretase, an enzyme critical to plaque formation.

“No one wants to talk about their secretase inhibitor research,” Dr. Thies said. “One of the amazing things about Elan is that the company did want to talk about their research, when not talking about it probably would have given them a competitive advantage.”

At this point, however, Elan is not letting AN-1792's inventor, Dr. Schenk, make any further comments on the vaccine or ongoing trials. Their public relations officials will say only that approximately 100 patients have taken the vaccine so far, and that AN-1792 apparently is safe. The company is still analyzing data from the phase I clinical safety trials, they said, and a larger safety trial will begin no earlier than late this year.

“It is a very, very optimistic time for AD research and in particular for the therapeutic phase of research,” Dr. Rosenberg observed. “Two or three years ago, this would not even have been a subject of discussion.”

©2001 American Academy of Neurology